PhantomJS

 

 

 

 

 

 

 

 

 

#!/usr/bin/env python
# -*- encoding: utf-8 -*-
# Created on 2015-08-03 14:33:08
# Project: nimabi

from pyspider.libs.base_handler import *


class Handler(BaseHandler):
    crawl_config = {
    }

    @every(minutes=24 * 60)
    def on_start(self):
        self.crawl(‘http://wanimal1983.tumblr.com/‘, callback=self.index_page)

    @config(age=10 * 24 * 60 * 60)
    def index_page(self, response):
        for each in response.doc(‘a[href^="http"]‘).items():
            self.crawl(each.attr.href, callback=self.detail_page)
    def list_page(self, response):
        for each in response.doc(‘.photoset-grid > img‘).items():
       
             self.crawl(each.attr.href, callback=self.detail_page)
    def detail_page(self, response):
        return {
            "url": response.url,
            "title": response.doc(‘title‘).text(),
            "rating": response.doc(‘.photoset-grid > img‘).text(),
            "imgs": [x.attr.src for x in response.doc(‘.photoset-grid img‘).items()]
        }